michael@0: /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
michael@0: /* This Source Code Form is subject to the terms of the Mozilla Public
michael@0:  * License, v. 2.0. If a copy of the MPL was not distributed with this
michael@0:  * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
michael@0: 
michael@0: #include "TestHarness.h"
michael@0: 
michael@0: nsresult TestGoodSurrogatePair()
michael@0: {
michael@0:   // When this string is decoded, the surrogate pair is U+10302 and the rest of
michael@0:   // the string is specified by indexes 2 onward.
michael@0:   const char16_t goodPairData[] = {  0xD800, 0xDF02, 0x65, 0x78, 0x0 };
michael@0:   nsDependentString goodPair16(goodPairData);
michael@0: 
michael@0:   uint32_t byteCount = 0;
michael@0:   char* goodPair8 = ToNewUTF8String(goodPair16, &byteCount);
michael@0:   if (!goodPair8)
michael@0:   {
michael@0:     fail("out of memory creating goodPair8");
michael@0:     return NS_ERROR_OUT_OF_MEMORY;
michael@0:   }
michael@0: 
michael@0:   if (byteCount != 6)
michael@0:   {
michael@0:     fail("wrong number of bytes; expected 6, got %lu", byteCount);
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   const unsigned char expected8[] =
michael@0:     { 0xF0, 0x90, 0x8C, 0x82, 0x65, 0x78, 0x0 };
michael@0:   if (0 != memcmp(expected8, goodPair8, sizeof(expected8)))
michael@0:   {
michael@0:     fail("wrong translation to UTF8");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   // This takes a different code path from the above, so test it to make sure
michael@0:   // the UTF-16 enumeration remains in sync with the UTF-8 enumeration.
michael@0:   nsDependentCString expected((const char*)expected8);
michael@0:   if (0 != CompareUTF8toUTF16(expected, goodPair16))
michael@0:   {
michael@0:     fail("bad comparison between UTF-8 and equivalent UTF-16");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   NS_Free(goodPair8);
michael@0: 
michael@0:   passed("TestGoodSurrogatePair");
michael@0:   return NS_OK;
michael@0: }
michael@0: 
michael@0: nsresult TestBackwardsSurrogatePair()
michael@0: {
michael@0:   // When this string is decoded, the two surrogates are wrongly ordered and
michael@0:   // must each be interpreted as U+FFFD.
michael@0:   const char16_t backwardsPairData[] = { 0xDDDD, 0xD863, 0x65, 0x78, 0x0 };
michael@0:   nsDependentString backwardsPair16(backwardsPairData);
michael@0: 
michael@0:   uint32_t byteCount = 0;
michael@0:   char* backwardsPair8 = ToNewUTF8String(backwardsPair16, &byteCount);
michael@0:   if (!backwardsPair8)
michael@0:   {
michael@0:     fail("out of memory creating backwardsPair8");
michael@0:     return NS_ERROR_OUT_OF_MEMORY;
michael@0:   }
michael@0: 
michael@0:   if (byteCount != 8)
michael@0:   {
michael@0:     fail("wrong number of bytes; expected 8, got %lu", byteCount);
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   const unsigned char expected8[] =
michael@0:     { 0xEF, 0xBF, 0xBD, 0xEF, 0xBF, 0xBD, 0x65, 0x78, 0x0 };
michael@0:   if (0 != memcmp(expected8, backwardsPair8, sizeof(expected8)))
michael@0:   {
michael@0:     fail("wrong translation to UTF8");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   // This takes a different code path from the above, so test it to make sure
michael@0:   // the UTF-16 enumeration remains in sync with the UTF-8 enumeration.
michael@0:   nsDependentCString expected((const char*)expected8);
michael@0:   if (0 != CompareUTF8toUTF16(expected, backwardsPair16))
michael@0:   {
michael@0:     fail("bad comparison between UTF-8 and malformed but equivalent UTF-16");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   NS_Free(backwardsPair8);
michael@0: 
michael@0:   passed("TestBackwardsSurrogatePair");
michael@0:   return NS_OK;
michael@0: }
michael@0: 
michael@0: nsresult TestMalformedUTF16OrphanHighSurrogate()
michael@0: {
michael@0:   // When this string is decoded, the high surrogate should be replaced and the
michael@0:   // rest of the string is specified by indexes 1 onward.
michael@0:   const char16_t highSurrogateData[] = { 0xD863, 0x74, 0x65, 0x78, 0x74, 0x0 };
michael@0:   nsDependentString highSurrogate16(highSurrogateData);
michael@0: 
michael@0:   uint32_t byteCount = 0;
michael@0:   char* highSurrogate8 = ToNewUTF8String(highSurrogate16, &byteCount);
michael@0:   if (!highSurrogate8)
michael@0:   {
michael@0:     fail("out of memory creating highSurrogate8");
michael@0:     return NS_ERROR_OUT_OF_MEMORY;
michael@0:   }
michael@0: 
michael@0:   if (byteCount != 7)
michael@0:   {
michael@0:     fail("wrong number of bytes; expected 7, got %lu", byteCount);
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   const unsigned char expected8[] =
michael@0:     { 0xEF, 0xBF, 0xBD, 0x74, 0x65, 0x78, 0x74, 0x0 };
michael@0:   if (0 != memcmp(expected8, highSurrogate8, sizeof(expected8)))
michael@0:   {
michael@0:     fail("wrong translation to UTF8");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   // This takes a different code path from the above, so test it to make sure
michael@0:   // the UTF-16 enumeration remains in sync with the UTF-8 enumeration.
michael@0:   nsDependentCString expected((const char*)expected8);
michael@0:   if (0 != CompareUTF8toUTF16(expected, highSurrogate16))
michael@0:   {
michael@0:     fail("bad comparison between UTF-8 and malformed but equivalent UTF-16");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   NS_Free(highSurrogate8);
michael@0: 
michael@0:   passed("TestMalformedUTF16OrphanHighSurrogate");
michael@0:   return NS_OK;
michael@0: }
michael@0: 
michael@0: nsresult TestMalformedUTF16OrphanLowSurrogate()
michael@0: {
michael@0:   // When this string is decoded, the low surrogate should be replaced and the
michael@0:   // rest of the string is specified by indexes 1 onward.
michael@0:   const char16_t lowSurrogateData[] = { 0xDDDD, 0x74, 0x65, 0x78, 0x74, 0x0 };
michael@0:   nsDependentString lowSurrogate16(lowSurrogateData);
michael@0: 
michael@0:   uint32_t byteCount = 0;
michael@0:   char* lowSurrogate8 = ToNewUTF8String(lowSurrogate16, &byteCount);
michael@0:   if (!lowSurrogate8)
michael@0:   {
michael@0:     fail("out of memory creating lowSurrogate8");
michael@0:     return NS_ERROR_OUT_OF_MEMORY;
michael@0:   }
michael@0: 
michael@0:   if (byteCount != 7)
michael@0:   {
michael@0:     fail("wrong number of bytes; expected 7, got %lu", byteCount);
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   const unsigned char expected8[] =
michael@0:     { 0xEF, 0xBF, 0xBD, 0x74, 0x65, 0x78, 0x74, 0x0 };
michael@0:   if (0 != memcmp(expected8, lowSurrogate8, sizeof(expected8)))
michael@0:   {
michael@0:     fail("wrong translation to UTF8");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   // This takes a different code path from the above, so test it to make sure
michael@0:   // the UTF-16 enumeration remains in sync with the UTF-8 enumeration.
michael@0:   nsDependentCString expected((const char*)expected8);
michael@0:   if (0 != CompareUTF8toUTF16(expected, lowSurrogate16))
michael@0:   {
michael@0:     fail("bad comparison between UTF-8 and malformed but equivalent UTF-16");
michael@0:     return NS_ERROR_FAILURE;
michael@0:   }
michael@0: 
michael@0:   NS_Free(lowSurrogate8);
michael@0: 
michael@0:   passed("TestMalformedUTF16OrphanLowSurrogate");
michael@0:   return NS_OK;
michael@0: }
michael@0: 
michael@0: 
michael@0: int main(int argc, char** argv)
michael@0: {
michael@0:   ScopedXPCOM xpcom("TestEncoding");
michael@0:   if (xpcom.failed())
michael@0:     return 1;
michael@0: 
michael@0:   int rv = 0;
michael@0: 
michael@0:   if (NS_FAILED(TestGoodSurrogatePair()))
michael@0:     rv = 1;
michael@0:   if (NS_FAILED(TestBackwardsSurrogatePair()))
michael@0:     rv = 1;
michael@0:   if (NS_FAILED(TestMalformedUTF16OrphanHighSurrogate()))
michael@0:     rv = 1;
michael@0:   if (NS_FAILED(TestMalformedUTF16OrphanLowSurrogate()))
michael@0:     rv = 1;
michael@0: 
michael@0:   return rv;
michael@0: }